The broad, long-term objective of this application is to determine the impact of oculomotor fatigue on diagnostic accuracy during the reading of digitally displayed radiological examinations.This will be accomplished with three specific aims: (1) test whether fatigue reduces detection accuracy and visual search efficiency in static image datasets, (2) test whether fatigue reduces detection accuracy and visual search efficiency in dynamic image datasets, and (3) discover the levels of visual fatigue that are experienced in a variety of situations faced by radiologists in daily practice. Current radiology practice relies increasingly on advanced imaging technology. Not only has this improved access to sub-specialists, yielding greater efficiency and productivity, it has also increased the radiologists' workload significantly. Radiologists now interpret more imaging studies, each containing many more images. This increase in interpretation workload lengthens the time they must spend viewing digital displays and this may adversely affect diagnostic performance. Extensive human factors engineering research has demonstrated that prolonged use of visual display terminals yields oculomotor fatigue, which in turn reduces performance. No research on eyestrain in radiology is available, but a preliminary study demonstrates that radiologists report increasingly severe symptoms of eyestrain, including blurred vision and difficulty focusing, as they read more imaging studies. We propose to evaluate measurable parameters of oculomotor fatigue such as accommodation, vergence, and visual search efficiency and relate them to diagnostic interpretation performance. Diagnostic accuracy will be analyzed with receiver operating characteristic (ROC) methods. Visual search efficiency will be studied by measuring saccade length, decision dwell, and type of search pattern. At the end of this project, we will understand the influence of oculomotor fatigue on diagnostic accuracy and interpretation efficiency. Only by developing a better understanding of the nature of observer error, may we discover effective approaches to reducing the errors.